---
title: "Yuxi Liu"
type: "entity"
entity_kind: "person"
status: "hub"
canonical_name: "Yuxi Liu"
aliases: ["Yuxi on the Wired"]
first_seen: "2026-07-06T00:00:00.000Z"
writer_model: "claude-sonnet-5"
connects_to: ["backpropagation","history of machine learning","Paul Werbos","citation history","Perceptrons (Minsky & Papert)"]
seek_code_commit: "17d9798"
---


Independent historian-blogger ("Yuxi on the Wired") whose essays reconstruct the citation and priority history of early neural-network research through direct archival and oral-history legwork. Matters to this vault as its single most-relied-on Tier 2 secondary voice on backpropagation's origins: his essays supply verified-verbatim quotes and bibliographic detail that anchor a large fraction of the vault's backprop-priority cluster, including the *Perceptrons* "sterile" quote, the Linnainmaa citation-count myth, and the RHW 1986 reference-list read.

His essay "The Backstory of Backpropagation" (2023, updated 2024) is the vault's route to Laveen Kanal's 1993 recollection about Werbos's obscure report and to Rumelhart's own stated reason for RHW's short 1986 reference list ("he didn't cite previous work because they were so obscure").

## References
- [[claim-rhw-1986-reference-list-four-works]] · [[claim-werbos-attributes-early-obscurity-to-publishing-strategy-not-suppression]] · [[claim-pollack-1989-sterile-quote-page-231]] · [[claim-perceptrons-multilayer-sterile-was-conjecture]] · [[claim-widrow-abandoned-multilayer-training-until-1985-backprop]] · [[myth-linnainmaa-uncited-before-2010s]] · [[claim-linnainmaa-priority-not-paternity]] · [[myth-perceptrons-book-killed-connectionism]] · [[claim-werbos-backprop-from-freud-own-account]]
- Captures: 2026-06-28 (RHW reference-list read), 2026-08-17-what-genuinely-connects-the-1986-entity-backpropagationbackpropagation-nature